8 ms·

"'Why? Because the Ph.D. system grinds people down'? Yes, and it’s completely inappropriate for what most people need. It was designed for German academics in

by gech 10y ago

"'Why? Because the Ph.D. system grinds people down'? Yes, and it’s completely inappropriate for what most people need. It was designed for German academics in the 19th century and it was fine for that. But for any other kind of life, it’s totally wrong. It takes far too long. It forces you to pretend to be a researcher when most people don’t want to be researchers. It’s become a union card and I think it’s highly disruptive, particularly bad for women. For women to waste five or 10 years of their lives is more of a disaster than it is for men." Anyone with relation to PhD programs care to reflect on this?

ozborn 10y ago
I would agree - I just finished my PhD in October. I did my PhD (Computer Science) while working full-time as a research associate and I was objectively less productive as a scientist while doing my PhD. I did it for the "union card" and I am now applying for faculty positions that would otherwise not have considered me without the PhD. It also agree it is very disruptive to family life and I think it not only drives women away from pursuing such a degree - it can in some cases contribute to women having fewer kids than they would like and/or increases the odds of them pursuing expensive fertility treatments later in life.
- coolride 10y ago
  Is it required for students applying to PhD programs in CS to have a very specific problem in mind before they can apply/be accepted to graduate school? Or do professors work with you to find a problem? Heard I absolutely cannot get into graduate school unless I have a very specific use case of a research problem I want to solve. Thought this would be difficult given my lack of knowledge in problems 'worth solving' before graduate studies. I do have a field of interest I want to go into and a broad problem I was to solve, but heard I need a very, very specific problem to propose before I can apply to graduate school.
scott_s 10y ago
I have a PhD in computer science. I work as a Research Staff Member at IBM Research. I mostly agree. One of the problems with the PhD system is that individuals have to carve out something that is theirs, they own it. That becomes their dissertation. It doesn't sound terrible said like that, but the most productive work I've ever done was after I graduated and got a job. A small group of people with equal ownership of a problem and solution can be so much more productive. The requirement that a student defend a body of work that is theirs disincentives them from doing, what I think, is the most productive kind of work that leads to the best research outcomes. You can work in productive groups in grad school, but you still need to find that part of the work that you can claim ownership of. Because if you don't, you won't graduate. This means that inevitably, you're going to have to think in terms of "what do I need to do to get the hell out of here" rather than "what's the best way we can solve this problem."
- Retric 10y ago
  This depends on the program, some are happy to use a set of joint publications as a defacto dissertation. Basically, several smaller chunks are used to represent a larger body of work, which allows for closer collaboration.
  - physicsyogi 10y ago
    When I finished my physics Ph.D. in 2005, my advisor showed me his dissertation. It was literally five papers with a cover around it. He graduated in 1985.
- Xcelerate 10y ago
  > I work as a Research Staff Member at IBM Research Funny you mention that. I recently finished my PhD and noticed a posting for a Research Staff Member position in quantum computing at IBM Research, and working on quantum computing research (or fundamental AI research) would be my ideal career. Normally, I submit my résumé and a cover letter to job listings like this, but I've found that it takes a long time to hear back (even when the company ultimately decides to offer me an interview). As someone who works at IBM yourself and has gone through the process, would you have any tips on improving my chances at being considered as a candidate? I haven't applied yet, but I'm just curious if there's a better route to piquing IBM's interest than submitting a résumé and waiting.
Shanea93 10y ago
Sorry, would you mind explaining why you consider it worse for women to "waste five or ten years of their lives"?
- jjaredsimpson 10y ago
  If you read the article you'd see the following interview question has the answer. > Because they might have family responsibilities as well? >> Yes, because there’s a biological clock. It’s ticking much more for them.
- JadeNB 10y ago
  It's a quote, not necessarily gech's opinion. In fact, https://news.ycombinator.com/item?id=13138792 https://news.ycombinator.com/item?id=13138792 explicitly asks for reflection on it.
kkylin 10y ago
For background, I have a PhD in applied/computational math, and am working in academia, though at an earlier point in my life I did contemplate going into the (software) industry. For whatever it's worth, I'm mostly happy with the choices I made, including getting a PhD in math. That said, I agree with the sentiment that a traditional PhD is not quite what many (I'm not sure about "most") people need. It makes sense to me to have a parallel system for providing training beyond a master's but isn't a traditional PhD. But working out what such a program would entail requires figuring out a lot of practical details, and it may be some time before such things become commonplace, if ever.
- madengr 10y ago
  What about a D.Eng? Is there an equivalent in other fields? https://en.m.wikipedia.org/wiki/Doctor_of_Engineering https://en.m.wikipedia.org/wiki/Doctor_of_Engineering
  - kkylin 10y ago
    I don't think it is all that common in engineering programs here in the US, though not being an engineer myself, I am really speaking out of ignorance here. I don't know if there is something equivalent in the sciences -- there are ScDs (https://en.wikipedia.org/wiki/Doctor_of_Science https://en.wikipedia.org/wiki/Doctor_of_Science) but as far as I can tell, it is equivalent to a PhD in all but name. Also, based on the description on Wikipedia, it seems like a D.Eng. is basically like a PhD except that the research problem can be more "applied" (not sure what this means -- it can mean very different things in different fields). In particular, a student still has to pass courses and exams, and in the end work on a piece of individual research as their dissertation.
BeetleB 10y ago
The "it takes far too long" for at least engineering PhDs is a very recent phenomenon. In the 90's, 3 years was the norm (after a MS). As for union card: Somewhat true. Varies depending on the discipline. As another commenter pointed out, though: With the right advisor, institution and topic, there can still be many pluses to make it worthwhile.
- semi-extrinsic 10y ago
  I completed my PhD this year after 3 years 4 months. In most European universities, 3-4 years is the norm, and you usually only have funding for that long anyway. Edit: we had one kid before starting the PhD and then another one half way through. Family and PhD is not unmixable. In fact, having a hard limit of 8 hours at the office per day surely helped me be more effective (I was responsible for dropping off and picking up from kindergarten every day).
  - stablemap 10y ago
    In math my impression from friends in the UK is that one is expected there to obtain an MS before entering a PhD program and that there are no teaching responsibilities. That definitely changes the numbers.
kurthr 10y ago
I feel like a lot of people go into a PhD program for mediocre reasons. Many of them would probably be better off doing a 2 year Masters program to learn 2 important things: 1) How to do research and run experiments (where you don't already know the answer). 2) How much they actually don't know about a very deep field (and how to study independently). The extra 4 years of a PhD, as I learned, were to make me an expert in a piece of the field with enough experience writing papers, meeting with colleagues, and designing my own research (not just next-step experiments), and writing funding proposals, that I could teach in the field. Like most others I chose not to become a professor... it's a very competitive and odd position. Part of the issue is the pressures that have been put on PhD programs. The fight for funding, papers, lab space politics, and relevance seem to have become all consuming as the number of students have grown and positions have remained fixed (filled and tenured). The days when you could follow a reasonable path into academia and Professorship without a lengthy post-doc are gone in most fields at most Universities. Professors have become unhappy managers of giant groups cranking out complex papers based on a huge apparatus (equipment, code, or lab) in order to get tenure and keep their position. CS and Math have a real advantage here as the tools for the job don't usually require as much yearly funding. Perhaps all this is fine for Particle Physics, but it feels like more and more of the positions at research universities are run this way... and it's not conducive to how many people want to earn their Doctorate... which is often just a badge of honor in the Business world anyway... so maybe we shouldn't try to push so many people through an academic mold so that they can have a different set of letters next to their name.

by gech 10y ago

ozborn 10y ago

I would agree - I just finished my PhD in October. I did my PhD (Computer Science) while working full-time as a research associate and I was objectively less productive as a scientist while doing my PhD. I did it for the "union card" and I am now applying for faculty positions that would otherwise not have considered me without the PhD. It also agree it is very disruptive to family life and I think it not only drives women away from pursuing such a degree - it can in some cases contribute to women having fewer kids than they would like and/or increases the odds of them pursuing expensive fertility treatments later in life.

coolride 10y ago
Is it required for students applying to PhD programs in CS to have a very specific problem in mind before they can apply/be accepted to graduate school? Or do professors work with you to find a problem? Heard I absolutely cannot get into graduate school unless I have a very specific use case of a research problem I want to solve. Thought this would be difficult given my lack of knowledge in problems 'worth solving' before graduate studies. I do have a field of interest I want to go into and a broad problem I was to solve, but heard I need a very, very specific problem to propose before I can apply to graduate school.

scott_s 10y ago

I have a PhD in computer science. I work as a Research Staff Member at IBM Research. I mostly agree. One of the problems with the PhD system is that individuals have to carve out something that is theirs, they own it. That becomes their dissertation. It doesn't sound terrible said like that, but the most productive work I've ever done was after I graduated and got a job. A small group of people with equal ownership of a problem and solution can be so much more productive. The requirement that a student defend a body of work that is theirs disincentives them from doing, what I think, is the most productive kind of work that leads to the best research outcomes. You can work in productive groups in grad school, but you still need to find that part of the work that you can claim ownership of. Because if you don't, you won't graduate. This means that inevitably, you're going to have to think in terms of "what do I need to do to get the hell out of here" rather than "what's the best way we can solve this problem."

Retric 10y ago
This depends on the program, some are happy to use a set of joint publications as a defacto dissertation. Basically, several smaller chunks are used to represent a larger body of work, which allows for closer collaboration.
- physicsyogi 10y ago
  When I finished my physics Ph.D. in 2005, my advisor showed me his dissertation. It was literally five papers with a cover around it. He graduated in 1985.
Xcelerate 10y ago
> I work as a Research Staff Member at IBM Research Funny you mention that. I recently finished my PhD and noticed a posting for a Research Staff Member position in quantum computing at IBM Research, and working on quantum computing research (or fundamental AI research) would be my ideal career. Normally, I submit my résumé and a cover letter to job listings like this, but I've found that it takes a long time to hear back (even when the company ultimately decides to offer me an interview). As someone who works at IBM yourself and has gone through the process, would you have any tips on improving my chances at being considered as a candidate? I haven't applied yet, but I'm just curious if there's a better route to piquing IBM's interest than submitting a résumé and waiting.

Shanea93 10y ago

Sorry, would you mind explaining why you consider it worse for women to "waste five or ten years of their lives"?

jjaredsimpson 10y ago
If you read the article you'd see the following interview question has the answer. > Because they might have family responsibilities as well? >> Yes, because there’s a biological clock. It’s ticking much more for them.
JadeNB 10y ago
It's a quote, not necessarily gech's opinion. In fact, https://news.ycombinator.com/item?id=13138792 https://news.ycombinator.com/item?id=13138792 explicitly asks for reflection on it.

kkylin 10y ago

For background, I have a PhD in applied/computational math, and am working in academia, though at an earlier point in my life I did contemplate going into the (software) industry. For whatever it's worth, I'm mostly happy with the choices I made, including getting a PhD in math. That said, I agree with the sentiment that a traditional PhD is not quite what many (I'm not sure about "most") people need. It makes sense to me to have a parallel system for providing training beyond a master's but isn't a traditional PhD. But working out what such a program would entail requires figuring out a lot of practical details, and it may be some time before such things become commonplace, if ever.

madengr 10y ago
What about a D.Eng? Is there an equivalent in other fields? https://en.m.wikipedia.org/wiki/Doctor_of_Engineering https://en.m.wikipedia.org/wiki/Doctor_of_Engineering
- kkylin 10y ago
  I don't think it is all that common in engineering programs here in the US, though not being an engineer myself, I am really speaking out of ignorance here. I don't know if there is something equivalent in the sciences -- there are ScDs (https://en.wikipedia.org/wiki/Doctor_of_Science https://en.wikipedia.org/wiki/Doctor_of_Science) but as far as I can tell, it is equivalent to a PhD in all but name. Also, based on the description on Wikipedia, it seems like a D.Eng. is basically like a PhD except that the research problem can be more "applied" (not sure what this means -- it can mean very different things in different fields). In particular, a student still has to pass courses and exams, and in the end work on a piece of individual research as their dissertation.

BeetleB 10y ago

The "it takes far too long" for at least engineering PhDs is a very recent phenomenon. In the 90's, 3 years was the norm (after a MS). As for union card: Somewhat true. Varies depending on the discipline. As another commenter pointed out, though: With the right advisor, institution and topic, there can still be many pluses to make it worthwhile.

semi-extrinsic 10y ago
I completed my PhD this year after 3 years 4 months. In most European universities, 3-4 years is the norm, and you usually only have funding for that long anyway. Edit: we had one kid before starting the PhD and then another one half way through. Family and PhD is not unmixable. In fact, having a hard limit of 8 hours at the office per day surely helped me be more effective (I was responsible for dropping off and picking up from kindergarten every day).
- stablemap 10y ago
  In math my impression from friends in the UK is that one is expected there to obtain an MS before entering a PhD program and that there are no teaching responsibilities. That definitely changes the numbers.

kurthr 10y ago

I feel like a lot of people go into a PhD program for mediocre reasons. Many of them would probably be better off doing a 2 year Masters program to learn 2 important things: 1) How to do research and run experiments (where you don't already know the answer). 2) How much they actually don't know about a very deep field (and how to study independently). The extra 4 years of a PhD, as I learned, were to make me an expert in a piece of the field with enough experience writing papers, meeting with colleagues, and designing my own research (not just next-step experiments), and writing funding proposals, that I could teach in the field. Like most others I chose not to become a professor... it's a very competitive and odd position. Part of the issue is the pressures that have been put on PhD programs. The fight for funding, papers, lab space politics, and relevance seem to have become all consuming as the number of students have grown and positions have remained fixed (filled and tenured). The days when you could follow a reasonable path into academia and Professorship without a lengthy post-doc are gone in most fields at most Universities. Professors have become unhappy managers of giant groups cranking out complex papers based on a huge apparatus (equipment, code, or lab) in order to get tenure and keep their position. CS and Math have a real advantage here as the tools for the job don't usually require as much yearly funding. Perhaps all this is fine for Particle Physics, but it feels like more and more of the positions at research universities are run this way... and it's not conducive to how many people want to earn their Doctorate... which is often just a badge of honor in the Business world anyway... so maybe we shouldn't try to push so many people through an academic mold so that they can have a different set of letters next to their name.